Article pubs.acs.org/ac
Speciation of Nitrogen-Bearing Species Using Negative and Positive Secondary Ion Spectra with Nano Secondary Ion Mass Spectrometry Kexue Li,§ Baerbel Sinha,*,§,# and Peter Hoppe§ §
Particle Chemistry Department, Max Planck Institute for Chemistry, 55128 Mainz, Germany Department of Earth and Environmental Sciences, Indian Institute of Science Education and Research Mohali, Sector 81 SAS Nagar, Manauli P.O., Punjab 140306, India
#
S Supporting Information *
ABSTRACT: In this study, we demonstrate that Nano Secondary Ion Mass Spectrometry (NanoSIMS) can be used to differentiate different nitrogen-containing species commonly observed in atmospheric aerosol particles with micrometer or submicrometer spatial resolution, on the basis of the relative intensity of secondary ion signals, both in negative and positive secondary ion mode, without the need to chemically or physically separate the samples. Compounds tested include nitrate, nitrite, ammonium salts, urea, amino acids, sugars, organic acids, amides, triazine, imidazole, protein, and biological tissue. We show that NO2− secondary ions are unique to the decomposition of nitrate and nitrite salts, whereas NH4+ secondary ions are unique to samples containing ammonium ions, with low signal intensities observed from amino groups but none from biological tissue. CN− signals are obtained from all nitrogen-bearing compounds, but relative signal intensities are the highest for organic nitrogen-containing compounds. We demonstrate that quantitative determination of the elemental fractions of carbon, oxygen, and nitrate in nanometer-sized aerosol samples using normalized secondary ion intensities is possible. We further demonstrate that stable isotope ratios measured on inhouse standards of unknown isotopic composition using the 12C15N−/12C14N− ratio (all nitrogen-containing species), the 15 16 N O2−/14N16O2− ratio (nitrate and nitrite species), and the 15NH4+/14NH4+ ratio (ammonium salts, amino acids, and urea) are stable and sufficiently precise for nitrogen isotope analysis.
T
analysis has the potential to overcome these restrictions in the interpretation of results. Second, separation of different inorganic and organic nitrogen-bearing species to obtain species-specific isotopic signatures represents an analytical challenge. Nr includes contributions from nitrate salts, ammonium salts, and organic nitrogen. All these species have different sources and are cycled via different reaction pathways. The sealed tube combustion method provides a single isotope signature for the total nitrogen in the sample,8,9 which is challenging to interpret. Separation of ammonium and nitrate ions via distillation or ion exchange chromatography involves an intensive, time-consuming sample preparation and requires a milligram quantity of sample material.10−13 The denitrifier method, which converts nitrate and nitrite to N2O (gas), is fast and requires only small samples, but recently, its specificity has been questioned.14 Moreover, the correction of the contribution of the massindependent oxygen isotope anomaly15 to m/z 45 and 46, which varies as a function of the oxidation pathway that
he atmospheric cycling of Nr (reactive nitrogen) is a focus of both scientific and policy concern, because of the importance of Nr in controlling the formation of tropospheric ozone. Atmospheric Nr also plays a role in the formation of particulate matter and its long-range transport modifies the global carbon cycle by providing nitrogen fertilization to remote ecosystems.1,2 Measurements of stable nitrogen isotope ratios offer a means of discriminating sources of nitrogen and reactions involved in Nr cycling via their specific isotopic signatures.3−7 The potential of using the stable nitrogen isotopic signatures of airborne particles to deepen our understanding of the sources, reactivity, transport, and removal of atmospheric nitrogen species has been recognized for a long time.4 Its application to real world samples, however, has suffered due to multiple analytical challenges. First, many conventional methods require large bulk samples. The analysis of species-specific nitrogen isotopic signatures of nitrate aerosol via the microbial denitrification, the most sensitive technique currently in use, typically requires several μg of material (20−60 nmol NO3−), which corresponds to 107 accumulation mode aerosol particles.4,8 Since atmospheric particles are chemically and morphologically heterogeneous, such an average signature is of limited use. Single particle © XXXX American Chemical Society
Received: December 15, 2015 Accepted: February 8, 2016
A
DOI: 10.1021/acs.analchem.5b04740 Anal. Chem. XXXX, XXX, XXX−XXX
Article
Analytical Chemistry
individually. The NanoSIMS 50 (basic configuration) can simultaneously measure up to five different masses via a multicollection system with high precision. More detailed information about the NanoSIMS 50 can be found elsewhere,31 and here we provide only a brief introduction of the settings used in this study. The NanoSIMS 50 has two primary ion sources: the CsCO3 source produces a beam of positively charged cesium primary ions (Cs+) and the duoplasmatron source, together with a Wien filter, provides a beam of negatively charged oxygen ions (O−). Primary ions are accelerated by a potential of ±8 kV in the source, to a potential of ±8 kV at the surface of the sample, thus impacting the sample with an energy of 16 keV.31 Before hitting the sample surface, the primary beam is trimmed by an aperture of variable size D1. Then it is focused by the EOP lens to its final spot size on the sample surface. Negative secondary ions 12C−, 12C14N−, 16O−, 14N16Ox− (x = 1, 2, 3) were measured with the Cs+ primary ion source. Aperture D1 was set to D1-2 (300 μm), which gave a primary ion beam current of ∼4 pA. Entrance slit and aperture slit were set to 20 × 140 μm2 (ES-4) and 150 × 150 μm2 (AS-3), respectively. The energy bandwidth was set to 20 eV. Positive secondary ions NHx+ (x = 1, 2, 3, 4) were measured with the duoplasmatron primary ion source. Aperture D1 was set to D1-2 (300 μm), which gave a primary ion beam current of 40 pA. Entrance slit and aperture slit were set to 20 × 140 μm (ES-4) and 80 × 80 μm (AS-4), respectively. The energy bandwidth was set to 20 eV. Description of Standards. A set of in-house standards of different nitrogen-bearing species listed in supplementary Table S1 were used to demonstrate how different nitrogen-bearing species can be differentiated from each other and from other organic compounds, which do not contain nitrogen using their secondary ion signals. These standards included four amino acids (alanine, serine, aspartic acid, cysteine), one protein (BSA), two organic acids (citric acid, adipic acid), two sugars (glucose, glucosamine HCl), two nitrates (sodium nitrate, potassium nitrate), one nitrite (sodium nitrite), two sulfates (ammonium sulfate, barium sulfate), and other organic compounds (ammonium oxalate, urea, N-1 adamanty1 urea, Tri(2-pyridyl)s-triazine, 1,8 naphtalimide, dicyandiamide, benzylimidazole, and polycarbonate). Preparation of Solid Samples. Agglomerates of grains of the in-house standards were pressed into ultraclean gold foils using the methodology described in Winterholler et al.19 Each sample holder contains several different standards. After pressing the samples into the gold foils, grain agglomerates measured several square millimeters for each standard. Subsequently, the mount was gold coated (Sputter coater Baltec SCD-050, sputter time = 120 s, sputtering current = 60 mA) to ensure sufficient surface conductivity of these larger area samples. The gold-coat thickness is about 70 nm. Preparation of Aerosol-Like Mixture Samples. Aerosollike mixture samples were produced using an in-house build atomizer from solutions containing 20g of solute dissolved in 1 L of Milli-Q water (Mili-Q produced by PURELAB Option-Q and recleaned by PURELAB UHQ before use). The aerosol was dried and collected on silicon wafers, mounted in a single stage impactor. The setup used to produce aerosol samples is shown in Figure 1. Aerosol particles with different N/C ratio and an aerodynamic diameter of ∼2 μm were prepared by mixing sodium nitrate and glucose and potassium nitrate and glucose, as detailed in supplementary Table S2.
produced the nitrate aerosol, often relies on unverified assumptions.4,14 Even on a single particle scale, different sources of Nr are frequently found internally mixed in the same particle. Therefore, the isotopic signature of the total nitrogen present in an individual particle, which can be easily obtained using the CN− molecular ion signal, is of limited use in understanding atmospheric nitrogen sources and the isotopic fingerprints of nitrogen cycling.16 It has recently been shown that speciesspecific nitrogen isotope analysis by NanoSIMS on bulk nitrate using the NO2− molecular ion is sufficiently accurate and precise for its potential application in the field of environmental sciences. However, for its widespread use, both the chemical composition and isotopic signature of individual aerosol particles need to be determined simultaneously, and the mixing state of the particles needs to be resolved.17 NanoSIMS has proven to be useful for studying isotopic signatures of individual atmospheric particles,18−20 and several previous studies demonstrated that the intensity of secondary ions in negative ion mode and that their ratios can be successfully used to classify aerosol particles into broad classes without a corresponding microscopy image. The secondary ion signals of 16O−, 12C2−, 12C14N−, and 32S− were used to identify soot, coated soot, organic aerosol, inorganic aerosol, primary biological aerosol, and mineral dust particles from NanoSIMS secondary ion images,20,21 and 12C−, 16O−, 12C14N−, 32S−, and 35 − Cl were used to separate particles from background and pollution episodes.22 In this study, we demonstrate that NanoSIMS can be used to quantitatively determine the abundance of carbon, oxygen, and nitrate in aerosol samples and to qualitatively differentiate different nitrogen-containing species commonly observed in atmospheric aerosol particles on the basis of the relative intensity of secondary ion signals, both in negative and positive secondary ion mode, without the need to chemically or physically separate the samples. Nitrogen-bearing species that occur in atmospheric aerosol particles include inorganic species such as nitrates, ammoniumcontaining salts like ammonium sulfate, ammonium nitrate, and ammonium chloride, organonitrates,16,23 amides, 16,24,25 amines,26−28 and imidazoles.29,30 NanoSIMS can be used to differentiate between these species with the help of 12C−, 16O−, 12 14 − C N , and 14N16O2− secondary ion signals under Cs+ bombardment in negative ion mode and with the help of the intensity of the 14NH4+, 14NH3+, 14NH2+, and 14NH+ molecular ion signals under O− bombardment in positive secondary ion mode. The species-specific nitrogen isotopic composition of individual aerosol particles can be monitored using the 14 16 N O2− signal, which is specific to nitrate/nitrite, the 14 NH4+ signal that shows high signal intensities only in ammonium-ion-containing salts and the 12C14N− signal, which can potentially be used to study the isotopic composition of the total organic and inorganic nitrogen present.
■
EXPERIMENTAL SECTION Mass Spectrometry. Measurements were performed with the Cameca NanoSIMS 50 ion probe at the Max Planck Institute for Chemistry (MPIC) in Mainz. This type of secondary ion mass spectrometer is characterized by high lateral resolution (down to 50 nm for Cs+ primary ions and 200−300 nm for O− primary ions). Therefore, micrometer and submicron-sized nitrogen aerosol particles can be measured B
DOI: 10.1021/acs.analchem.5b04740 Anal. Chem. XXXX, XXX, XXX−XXX
Article
Analytical Chemistry
Figure 2. Relative intensities of secondary ions 12C14N−, 14N16O−, 14 16 N O2−, and 14N16O3− in sodium nitrate and sodium nitrite samples. The maximum signal 14N16O2− is set to 100 and the weaker signals are expressed in % of the strongest signal.
Figure 1. Setup used to prepare aerosol standards.
Preparation of Pure Aerosol Samples. The sample preparation has been described in Pöhlker et al.32 Briefly, pure compounds were dissolved in Milli-Q water (3 mM solution). The solution was converted into aerosol using an atomizer, dried, passed through a differential mobility analyzer (DMA), and deposited on sample substrates (Si3N4 windows) mounted in a single stage impactor. The aerosol thus produced had mobility diameters of 0.5 μm. Collection of Ambient Aerosol Samples. Ambient aerosol samples from the Amazon rain forest were deposited on Si3N4 windows mounted in a single stage impactor. Further details can be found in Pöhlker et al.32 Figures S4 and S5 show several ambient particles (PBA, potassium sulfate + SOA + POA, soot + salt coating and ammonium sulfate + SOA), which were imaged both using the scanning electron microscope (SEM) and the NanoSIMS. Nitrogen Isotope Ratio Calculation. Nitrogen (N) has two naturally occurring stable isotopes: one with a relative atomic mass of 14 (14N) and the other of 15 (15N). The most common, 14N, has an abundance in N2 gas of 99.63%.33,34 Nitrogen isotope measurements by SIMS are difficult because the positive secondary ion yield of N+ is very low, and negative secondary ions do not form at all.31,35 In this paper, we investigate the feasibility of species-specific nitrogen isotope analysis using three secondary ion ratios: 12C15N−/12C14N−, 15 16 N O2−/14N16O2−, and 15NH4+/14NH4+. Uncorrected isotope ratios (i.e., ratios not corrected for instrumental mass fractionation and not calibrated against an internationally accredited standard) were calculated as
signals.37,36 The higher NO3− ion signals of earlier studies can be attributed to the experimental conditions, which were optimized toward obtaining accurate chemical information (single-projectile impacts on pristine sample surfaces) but not targeted toward producing count rates high enough for simultaneous isotope analysis. As we use high primary ion densities of ∼1.5 × 109 Cs+ primary ions per μm2 of sample surface, a significant fraction of the primary ions impacts a sample surface already disturbed by previous collision cascades and ejects secondary ions for which one of the N−O bonds has already been broken. The ratio of 14N16O−/14N16O2− and 14N16O3−/14N16O2− are 12.8% and 11.4%, respectively, for sodium nitrate and 30.0% and ∼0%, respecitvely, for sodium nitrite. Very few counts of the 14N16O3− molecular ion were observed for sodium nitrite, which indicates that the 14N16O3− molecular ion is not formed in the reaction zone above the sample but only produced by direct sputtering of samples containing 14N16O3− ions. This is useful to separate nitrate and nitrite salts during NanoSIMS analysis. However, the 14N16O3− secondary ion is unsuitable for high-precision stable isotope analysis for two reasons. First, it is impossible to bring two detectors close enough together for simultaneous analysis of 14N16O3− and 15N16O3− in a NanoSIMS with basic configuration. Measurements in peak jumping mode are undesirable on aerosol particles as they carry a much larger measurement uncertainty. Second, the signal intensity is ∼1/9 lower than that of NO2− molecular ions. Considering the constraints imposed by the small sample size available in most aerosol particles, precise analysis will usually be difficult due to the low signal intensity. In order to measure the nitrogen isotope ratio with higher precision, the species with the larger secondary ion yield was chosen for measuring the speciesspecific nitrogen isotopic composition of nitrate-containing salts. The same species was also chosen as the most suitable species to detect nitrate/nitrite-containing salts in atmospheric aerosol samples. 14N16O3− can be measured on a third detector during isotope analysis to separate these two. Mass Spectra of Ammonium-Containing Inorganic Salts and Organic Amino Groups. Only few studies so far reported measuring ammonium-containing salts using SIMS. Previous researchers reported that ammonium salts can be identified by their positive secondary ion mass spectra which displays a cascade of NHx+ (x = 1, 2, 3, 4) with decreasing signal intensity for smaller x.24 They also reported that the same fragmentation pattern is not observed in organic amino compounds. However, both studies did not report detailed mass spectra. Figure 3 shows that we observe a cascade of secondary ions with 14NH4+ having the highest intensity compared to 14NH+, 14NH2+, and 14NH3+. The strongest signal
⎞ ⎛R δ15Nuncorrected = ⎜ X − 1⎟ × 1000(‰) ⎠ ⎝ RX
where RX represents 12C15N−/12C14N−, 15N16O2−/14N16O2−, or 15 NH4+/14NH4+, and RX the average of all measurements.
■
RESULTS AND DISCUSSION Mass Spectra of Nitrate and Nitrite Species. Benz et al. were the first to report that 16O2− and 14N16Ox− were the major ions observed while sputtering sodium nitrate with Ar+.24 The same study reported that these molecular ions were uniquely associated with inorganic nitrates and nitrite and could be used to separate these from organic nitrogen compounds and ammonium-containing inorganic salts. The decomposition patterns of nitrate and nitrite caused by primary ions in SIMS analysis was reported by Groenewold et al. and Van Stipdonk et al.36,37 Figure 2 shows the seconday ion intensity of 14 16 N Ox− measured with the NanoSIMS. The strongest signal 14 16 ( N O2−) has been set to 100 and the weaker signals are expressed in percent of the strongest signal. We find that the secondary ion 14N16O2− signal has the highest intensity, while previous studies reported much higher 14N16O3− secondary ion C
DOI: 10.1021/acs.analchem.5b04740 Anal. Chem. XXXX, XXX, XXX−XXX
Article
Analytical Chemistry
probably because they have the strongest chemical bonding between carbon and nitrogen in molecules. Zinner et al. also reported that different chemical bonding between carbon and nitrogen influences the CN− secondary ion intensity.35 Proteins and complex biological tissue exhibits lower 12C14N− and higher 12C− signal intensities. Inorganic compounds show high 16 − O secondary ion signals (sulfates, nitrates, nitrite). However, even the trace levels of carbon contamination in inorganic nitrogen-containing salts are sufficient to produce a CN− secondary ion signal which is potentially strong enough to allow isotope analysis. Similarly, nitrogen-free organic compounds (glucose, citric acid adipinic acid, and polycarbonate) display a strong 12C− negative secondary ion signal, but at the same time, trace-level nitrogen contamination picked up during the aerosol preparation results in high CN− signals in the glucose, citric acid, and adipinic acid aerosol samples. This problem is not observed in solid samples (polycarbonate). 16O− signals correspond with the abundance of oxygen in the samples. Analysis of Nitrogen-Containing Aerosol-Like Mixture Samples. Figure 5 shows 5 × 5 μm2 negative secondary ion images of aerosol-like mixture samples of KNO3 mixed with glucose at different mixing ratios. It can be seen that for samples containing ≥50% nitrate, the salt dries first, while the organic phase retains traces of the salt and displays delayed drying behavior.39 As a consequence, the crystal habitus of the inorganic core becomes more pronounced for mixtures with larger organic mass fraction. Pure salt particles crystallize rapidly with the habitus of a spherical water droplet upon drying. The organic phase and the salt are mixed throughout the full particle only when the salt content is low (50% nitrate. In samples with
